My far the most common type of spherical UAS is a balloon. Specifically a weather/scientific/"into space" balloon.
My #1 hypothesis would be a spherical balloon with some fairly heavy payload. It did not burst at altitude for some reason (underinflation, maybe deliberate), and instead ended up drifting down. It hit the water payload first, and the wind caused a rip in the ballon, instant deflation, sinks.
The key to attaining neutral buoyancy at a specific altitude is onboarding just enough gas to provide the required lift to ascend from the launch altitude, yet not so much that the balloon overshoots the target altitude. The balloon stops rising and becomes neutrally buoyant (if everything is done correctly) when the mass of the atmosphere displaced by the balloon and payload is equal to the mass of the gas inside the balloon plus the mass of the balloon and payload. The primary variable that makes neutral buoyancy possible is the elasticity of the balloon and the resulting increase of pressure inside the balloon as compared to the atmospheric pressure. This pressure increase is more significant at high altitudes thereby allowing the balloon to become neutrally buoyant. The trick is to give the balloon the "Goldilocks" amount of gas, as Dr. Maxham calls it: just enough to get it to the desired altitude, but not too much so that the balloon can't contain it when it gets there.
An issue with this is that UV radiation apparently degrades the ballon, causing it to pop before it sinks to the ground. However this might not be an issue if it was designed for low altitude flight.